The present invention provides a technique for separately encoding a plurality of multi-view images and additional information matching the multi-view images. A method for decoding for video data including combination information comprises the following steps of: a decoder receiving a plurality of images capturing different directions at the same time, feature information and mapping information with respect to each of the plurality of images; the decoder combining the plurality of images into one image by stitching the plurality of images based on the feature information; and the decoder converting one image into a 360-degree image based on the mapping information. The feature information which is a feature value extracted from image is used to combine the image capturing an adjacent region among the plurality of images. The 360-degree image provides the image at a specific viewpoint in a three-dimensional space according to a direction of a playing device.
展开▼
